The most important part of the SID API is the query endpoint, available at https://api.sid.ai/v1/users/me/query. This is the endpoint that allows you to query the user’s data that has been indexed by SID. In this section, we’ll discuss basic usage examples of the endpoint, and limitations that you should keep in mind when using it.

In its simplest form, a query is a JSON object with a single key, query, whose value is a string, for example:

{
  "query": "what does SID do?"
}

This query will return a list of documents that are relevant to the query, along with a score and certain metadata. An example response is shown below:

{
  "results": [
    {
      "score": 0.9470948562,
      "title": "SID Company Profile",
      "kind": "gdrive",
      "text": "SID is a company that makes software for searching and organizing information. ...",
    },
    {
      "score": 0.6341365791,
      "title": "Quarterly Results",
      "kind": "notion",
      "text": "In the last quarter, SID has integrated with multiple notable companies, including ...",
    },
    /* ... */
  ],
  "meta": {
    "sync_in_progress": false,
  }
}

Assuming the original query comes from a user, you could, for example, feed the results into an LLM to further refine them, and then display them to the user.

Tips for Querying

Querying the SID API is akin to using a search engine like Google. You can use natural language queries to get good results. For instance, a query like

What are the main features of SID, and how can they help me?

would yield relevant results. However, if you’re looking for something specific, using keywords can yield better results. For example, if you have a document titled “SID primary features”, then a query like

main features sid

would have a higher likelihood of finding this document than a more complex query in natural language.

API Reference

Learn more about the query endpoint

Was this page helpful?